Development of MATLAB® GUI for infant cry signal analysis using FFT / Suzilawati Muhamud @ Kayat

Crying is a universal act in infancy and an essential signal that activates care giving behavior. It is a normal, healthy means of expression and communication especially to newborn infants. At this stage, crying is the only way to communicate their needs. The cry signal has enormous potential diagnostic value. A very high-pitched cry tells that something may be wrong with the infant and this signal can be an early warning leads to further neurological testing. Infant cry analysis, reported in the literature, has been a topic for quite a number of researchers for the past 40 years. In previous works on the analysis of infant cry, it has been shown that there exist significant differences among the several types of crying, such as healthy, pain, hunger and pathological infant cry. In a recent investigation of infant cry analysis, Zaidi Shaari [3] attempted to analyze the infant cry signals using Short Time Fourier Transform (STFT). The goal of Zaidi Shaari was to differentiate between both pain and hunger cries by analyzing the distribution of energy in both frequency and time. The thesis is an enhancement of the previous work of Zaidi Shaari. The main objective of this thesis is to develop a Graphical User Interface (GUI) in MATLAB for infant cry signal analysis. The audio signals of the infant cry were obtained from a database available on the website (DISAT, Universita degli Studi di Milano-Bicocca, Milano, Italy) [7]. The signals were recorded from a total of six normal newborn babies of age four days regardless of gender. Two types of cry signals namely hunger and pain were being analyzed. FFT analysis and Hilbert transformation were used to extract information embedded in both cry signals Algorithms and programming utilizing MATLAB R2007b were written to search for the fundamental frequency of each cry signal and to identify the type of signal uniquely according to its fundamental frequency. The development; of GUI requires a good understanding and skills in programming the M-File code and it was the most challenging part for this study.